I’m looking for a PhD student to work on using Artificial Intelligence (AI) and Natural Language Generation (NLG) to help cancer patients who are managing their condition at home (FindAPhD). The student will be supervised by faculty at Aberdeen’s Medical School as well as by me. I think this is a very exciting PhD, but its also unusual in some ways, so I thought I’d explain more about it in a blog.

Problem

50% of people with cancer in England and Wales live for at least ten years after being diagnosed. During most of this time, they will be at home, not in the hospital. Such people need to manage and monitor their cancer while at home, and alert their healthcare team if their cancer is getting worse. Some people do this well, but others do not.

Apps can help people manage and monitor cancer at home. Existing apps focus on gathering data, communicating with the health service, and giving advice. This is definitely useful, but we think there is considerable potential in using AI and NLG technology to make such apps more effective, including communicating information in a more personalised, engaging, and understandable manner. Ie, this is an area where AI and NLG can make a real difference to people’s lives!

PhD

The goal of the PhD is to understand how AI and NLG can help, focusing on patients in Scotland with colorectal and prostrate cancer. It is a PhD in Medical Sciences, not Computer Science, which means that the focus is not just on technology. The PhD will include

• Reviewing current usage of AI/NLG in this space. How effective are current solutions in actually helping patients?
• Working with patients and healthcare providers to understand their problems and information/communication needs. What would actually help patients and healthcare providers?
• Building an app (possibly a chatbot) that supports patients and communicates key information, and also produces reports for healthcare professionals. The app should be usable by a wide range of patients, so there are HCI as well as AI/NLG challenges.
• Evaluation of the effectiveness of the app, with patients and healthcare professionals.

In other words, there is a lot more to this project than just building an AI system! We want to develop technology which can help solve a real-world problem, which means that requirements, evaluation, and usability are as important as clever AI technology.

Supervision

The PhD will be supervised by Dr Rosalind Adam and Prof Peter Murchie from Aberdeen’s Medical School, as well as by me. Dr Adam (who is an actual doctor as well as a PhD) has a 5-year fellowship to work on using eHealth to improve cancer survivorship; this PhD is basically funded from Dr Adam’s fellowship. Professor Murchie has been working on apps to support cancer survivors for many years, including clinical trials of such apps. I have been working on using NLG in healthcare contexts for decades, and there should be links between this project and research we are currently doing in the PhilHumans project.

Student

We’re looking for someone with a strong background or at least interest in both AI/Computing and in healthcare. The candidate also needs to be able work with patients and healthcare providers.

We also want candidates who are passionate on using AI and NLG to solve real-world problems. People whose main goal is to get top position on an ML/NLP leaderboard should probably not apply for this position.

The PhD studentship is fully funded for UK students. International students will need to pay the difference between international and UK tuition fees. For more information on how to apply and other practicalities, please see FindAPhD

Application deadline is 16 August.

I think this could be a really exciting opportunity for people with the above background!

PS

Just as a reminder, Alberto Bugarín Diz is looking for a PhD student to work on explaining Bayesian models who I will co-supervise.